Distributions
Gumbel Softmax
Introduced by Jang et al. in Categorical Reparameterization with Gumbel-SoftmaxGumbel-Softmax is a continuous distribution that has the property that it can be smoothly annealed into a categorical distribution, and whose parameter gradients can be easily computed via the reparameterization trick.
Source: Categorical Reparameterization with Gumbel-Softmax
Papers
| Paper | Code | Results | Date | Stars |
|---|
Tasks
| Task | Papers | Share |
|---|---|---|
| Clustering | 6 | 7.79% |
| Image Classification | 4 | 5.19% |
| Speech Recognition | 4 | 5.19% |
| General Classification | 4 | 5.19% |
| Reinforcement Learning (RL) | 3 | 3.90% |
| Quantization | 3 | 3.90% |
| Recommendation Systems | 3 | 3.90% |
| Diversity | 2 | 2.60% |
| Evolutionary Algorithms | 2 | 2.60% |
Usage Over Time
Components
| Component | Type |
|
|---|---|---|
| 🤖 No Components Found | You can add them if they exist; e.g. Mask R-CNN uses RoIAlign |
